xen: bug fixes for 3.18-rc1

- Fix regression in xen_clocksource_read() which caused all Xen guests
   to crash early in boot.
 - Several fixes for super rare race conditions in the p2m.
 - Assorted other minor fixes.
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.12 (GNU/Linux)
 
 iQEcBAABAgAGBQJUSh3nAAoJEFxbo/MsZsTRw6IH/imL2J++b8cafVvHjmVRt1T/
 P7KuFYPh/Tym+LISDBfk7MeOXZWsffvUDP653cGQiIMgmumEgVrU1+vR2Z0qRiRe
 95ZDIuQBmyGNBG9MiB0+zB7+STsvLECkPVWYDJCNbGVgrlHL6UHne06edrSpfr30
 13PyZeJAojezrt2hzLO43V7bu9acRmLo6WNdh6N2stfJv8QSQYSQO87baRdRB+rO
 I1r2jP7TJp9ZRtzSTsYLfpyhCGLcvXY58bci+Tz9x6xWMJ/HH5HvfJjxO17HzbdD
 2se6MKFVbOXT7DQK+BvQBDIO52t731DWZs4t7SJg24kDoINL7XiC/qSHC0vHJJM=
 =Cs0b
 -----END PGP SIGNATURE-----

Merge tag 'stable/for-linus-3.18-b-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip

Pull xen bug fixes from David Vrabel:

 - Fix regression in xen_clocksource_read() which caused all Xen guests
   to crash early in boot.
 - Several fixes for super rare race conditions in the p2m.
 - Assorted other minor fixes.

* tag 'stable/for-linus-3.18-b-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  xen/pci: Allocate memory for physdev_pci_device_add's optarr
  x86/xen: panic on bad Xen-provided memory map
  x86/xen: Fix incorrect per_cpu accessor in xen_clocksource_read()
  x86/xen: avoid race in p2m handling
  x86/xen: delay construction of mfn_list_list
  x86/xen: avoid writing to freed memory after race in p2m handling
  xen/balloon: Don't continue ballooning when BP_ECANCELED is encountered

arch/x86/xen/enlighten.c

@@ -1636,9 +1636,6 @@ asmlinkage __visible void __init xen_start_kernel(void)
 	xen_raw_console_write("mapping kernel into physical memory\n");
 	xen_setup_kernel_pagetable((pgd_t *)xen_start_info->pt_base, xen_start_info->nr_pages);
 
-	/* Allocate and initialize top and mid mfn levels for p2m structure */
-	xen_build_mfn_list_list();
-
 	/* keep using Xen gdt for now; no urgent need to change it */
 
 #ifdef CONFIG_X86_32

arch/x86/xen/mmu.c

@@ -1217,10 +1217,13 @@ static void __init xen_pagetable_p2m_copy(void)
 static void __init xen_pagetable_init(void)
 {
 	paging_init();
-	xen_setup_shared_info();
 #ifdef CONFIG_X86_64
 	xen_pagetable_p2m_copy();
 #endif
+	/* Allocate and initialize top and mid mfn levels for p2m structure */
+	xen_build_mfn_list_list();
+
+	xen_setup_shared_info();
 	xen_post_allocator_init();
 }
 static void xen_write_cr2(unsigned long cr2)

arch/x86/xen/p2m.c

@@ -163,6 +163,7 @@
 #include <linux/hash.h>
 #include <linux/sched.h>
 #include <linux/seq_file.h>
+#include <linux/bootmem.h>
 
 #include <asm/cache.h>
 #include <asm/setup.h>
@@ -181,21 +182,20 @@ static void __init m2p_override_init(void);
 
 unsigned long xen_max_p2m_pfn __read_mostly;
 
+static unsigned long *p2m_mid_missing_mfn;
+static unsigned long *p2m_top_mfn;
+static unsigned long **p2m_top_mfn_p;
+
 /* Placeholders for holes in the address space */
 static RESERVE_BRK_ARRAY(unsigned long, p2m_missing, P2M_PER_PAGE);
 static RESERVE_BRK_ARRAY(unsigned long *, p2m_mid_missing, P2M_MID_PER_PAGE);
-static RESERVE_BRK_ARRAY(unsigned long, p2m_mid_missing_mfn, P2M_MID_PER_PAGE);
 
 static RESERVE_BRK_ARRAY(unsigned long **, p2m_top, P2M_TOP_PER_PAGE);
-static RESERVE_BRK_ARRAY(unsigned long, p2m_top_mfn, P2M_TOP_PER_PAGE);
-static RESERVE_BRK_ARRAY(unsigned long *, p2m_top_mfn_p, P2M_TOP_PER_PAGE);
 
 static RESERVE_BRK_ARRAY(unsigned long, p2m_identity, P2M_PER_PAGE);
 static RESERVE_BRK_ARRAY(unsigned long *, p2m_mid_identity, P2M_MID_PER_PAGE);
-static RESERVE_BRK_ARRAY(unsigned long, p2m_mid_identity_mfn, P2M_MID_PER_PAGE);
 
 RESERVE_BRK(p2m_mid, PAGE_SIZE * (MAX_DOMAIN_PAGES / (P2M_PER_PAGE * P2M_MID_PER_PAGE)));
-RESERVE_BRK(p2m_mid_mfn, PAGE_SIZE * (MAX_DOMAIN_PAGES / (P2M_PER_PAGE * P2M_MID_PER_PAGE)));
 
 /* For each I/O range remapped we may lose up to two leaf pages for the boundary
  * violations and three mid pages to cover up to 3GB. With
@@ -272,11 +272,11 @@ static void p2m_init(unsigned long *p2m)
  * Build the parallel p2m_top_mfn and p2m_mid_mfn structures
  *
  * This is called both at boot time, and after resuming from suspend:
- * - At boot time we're called very early, and must use extend_brk()
+ * - At boot time we're called rather early, and must use alloc_bootmem*()
  *   to allocate memory.
  *
  * - After resume we're called from within stop_machine, but the mfn
- *   tree should alreay be completely allocated.
+ *   tree should already be completely allocated.
  */
 void __ref xen_build_mfn_list_list(void)
 {
@@ -287,20 +287,17 @@ void __ref xen_build_mfn_list_list(void)
 
 	/* Pre-initialize p2m_top_mfn to be completely missing */
 	if (p2m_top_mfn == NULL) {
-		p2m_mid_missing_mfn = extend_brk(PAGE_SIZE, PAGE_SIZE);
+		p2m_mid_missing_mfn = alloc_bootmem_align(PAGE_SIZE, PAGE_SIZE);
 		p2m_mid_mfn_init(p2m_mid_missing_mfn, p2m_missing);
-		p2m_mid_identity_mfn = extend_brk(PAGE_SIZE, PAGE_SIZE);
-		p2m_mid_mfn_init(p2m_mid_identity_mfn, p2m_identity);
 
-		p2m_top_mfn_p = extend_brk(PAGE_SIZE, PAGE_SIZE);
+		p2m_top_mfn_p = alloc_bootmem_align(PAGE_SIZE, PAGE_SIZE);
 		p2m_top_mfn_p_init(p2m_top_mfn_p);
 
-		p2m_top_mfn = extend_brk(PAGE_SIZE, PAGE_SIZE);
+		p2m_top_mfn = alloc_bootmem_align(PAGE_SIZE, PAGE_SIZE);
 		p2m_top_mfn_init(p2m_top_mfn);
 	} else {
 		/* Reinitialise, mfn's all change after migration */
 		p2m_mid_mfn_init(p2m_mid_missing_mfn, p2m_missing);
-		p2m_mid_mfn_init(p2m_mid_identity_mfn, p2m_identity);
 	}
 
 	for (pfn = 0; pfn < xen_max_p2m_pfn; pfn += P2M_PER_PAGE) {
@@ -328,10 +325,9 @@ void __ref xen_build_mfn_list_list(void)
 			/*
 			 * XXX boot-time only!  We should never find
 			 * missing parts of the mfn tree after
-			 * runtime.  extend_brk() will BUG if we call
-			 * it too late.
+			 * runtime.
 			 */
-			mid_mfn_p = extend_brk(PAGE_SIZE, PAGE_SIZE);
+			mid_mfn_p = alloc_bootmem_align(PAGE_SIZE, PAGE_SIZE);
 			p2m_mid_mfn_init(mid_mfn_p, p2m_missing);
 
 			p2m_top_mfn_p[topidx] = mid_mfn_p;
@@ -415,7 +411,6 @@ void __init xen_build_dynamic_phys_to_machine(void)
 	m2p_override_init();
 }
 #ifdef CONFIG_X86_64
-#include <linux/bootmem.h>
 unsigned long __init xen_revector_p2m_tree(void)
 {
 	unsigned long va_start;
@@ -477,7 +472,6 @@ unsigned long __init xen_revector_p2m_tree(void)
 
 			copy_page(new, mid_p);
 			p2m_top[topidx][mididx] = &mfn_list[pfn_free];
-			p2m_top_mfn_p[topidx][mididx] = virt_to_mfn(&mfn_list[pfn_free]);
 
 			pfn_free += P2M_PER_PAGE;
 
@@ -538,12 +532,13 @@ static bool alloc_p2m(unsigned long pfn)
 	unsigned topidx, mididx;
 	unsigned long ***top_p, **mid;
 	unsigned long *top_mfn_p, *mid_mfn;
+	unsigned long *p2m_orig;
 
 	topidx = p2m_top_index(pfn);
 	mididx = p2m_mid_index(pfn);
 
 	top_p = &p2m_top[topidx];
-	mid = *top_p;
+	mid = ACCESS_ONCE(*top_p);
 
 	if (mid == p2m_mid_missing) {
 		/* Mid level is missing, allocate a new one */
@@ -558,7 +553,7 @@ static bool alloc_p2m(unsigned long pfn)
 	}
 
 	top_mfn_p = &p2m_top_mfn[topidx];
-	mid_mfn = p2m_top_mfn_p[topidx];
+	mid_mfn = ACCESS_ONCE(p2m_top_mfn_p[topidx]);
 
 	BUG_ON(virt_to_mfn(mid_mfn) != *top_mfn_p);
 
@@ -566,6 +561,7 @@ static bool alloc_p2m(unsigned long pfn)
 		/* Separately check the mid mfn level */
 		unsigned long missing_mfn;
 		unsigned long mid_mfn_mfn;
+		unsigned long old_mfn;
 
 		mid_mfn = alloc_p2m_page();
 		if (!mid_mfn)
@@ -575,17 +571,19 @@ static bool alloc_p2m(unsigned long pfn)
 
 		missing_mfn = virt_to_mfn(p2m_mid_missing_mfn);
 		mid_mfn_mfn = virt_to_mfn(mid_mfn);
-		if (cmpxchg(top_mfn_p, missing_mfn, mid_mfn_mfn) != missing_mfn)
+		old_mfn = cmpxchg(top_mfn_p, missing_mfn, mid_mfn_mfn);
+		if (old_mfn != missing_mfn) {
 			free_p2m_page(mid_mfn);
-		else
+			mid_mfn = mfn_to_virt(old_mfn);
+		} else {
 			p2m_top_mfn_p[topidx] = mid_mfn;
+		}
 	}
 
-	if (p2m_top[topidx][mididx] == p2m_identity ||
-	    p2m_top[topidx][mididx] == p2m_missing) {
+	p2m_orig = ACCESS_ONCE(p2m_top[topidx][mididx]);
+	if (p2m_orig == p2m_identity || p2m_orig == p2m_missing) {
 		/* p2m leaf page is missing */
 		unsigned long *p2m;
-		unsigned long *p2m_orig = p2m_top[topidx][mididx];
 
 		p2m = alloc_p2m_page();
 		if (!p2m)
@@ -606,7 +604,6 @@ static bool __init early_alloc_p2m(unsigned long pfn, bool check_boundary)
 {
 	unsigned topidx, mididx, idx;
 	unsigned long *p2m;
-	unsigned long *mid_mfn_p;
 
 	topidx = p2m_top_index(pfn);
 	mididx = p2m_mid_index(pfn);
@@ -633,43 +630,21 @@ static bool __init early_alloc_p2m(unsigned long pfn, bool check_boundary)
 
 	p2m_top[topidx][mididx] = p2m;
 
-	/* For save/restore we need to MFN of the P2M saved */
-
-	mid_mfn_p = p2m_top_mfn_p[topidx];
-	WARN(mid_mfn_p[mididx] != virt_to_mfn(p2m_missing),
-		"P2M_TOP_P[%d][%d] != MFN of p2m_missing!\n",
-		topidx, mididx);
-	mid_mfn_p[mididx] = virt_to_mfn(p2m);
-
 	return true;
 }
 
 static bool __init early_alloc_p2m_middle(unsigned long pfn)
 {
 	unsigned topidx = p2m_top_index(pfn);
-	unsigned long *mid_mfn_p;
 	unsigned long **mid;
 
 	mid = p2m_top[topidx];
-	mid_mfn_p = p2m_top_mfn_p[topidx];
 	if (mid == p2m_mid_missing) {
 		mid = extend_brk(PAGE_SIZE, PAGE_SIZE);
 
 		p2m_mid_init(mid, p2m_missing);
 
 		p2m_top[topidx] = mid;
-
-		BUG_ON(mid_mfn_p != p2m_mid_missing_mfn);
-	}
-	/* And the save/restore P2M tables.. */
-	if (mid_mfn_p == p2m_mid_missing_mfn) {
-		mid_mfn_p = extend_brk(PAGE_SIZE, PAGE_SIZE);
-		p2m_mid_mfn_init(mid_mfn_p, p2m_missing);
-
-		p2m_top_mfn_p[topidx] = mid_mfn_p;
-		p2m_top_mfn[topidx] = virt_to_mfn(mid_mfn_p);
-		/* Note: we don't set mid_mfn_p[midix] here,
-		 * look in early_alloc_p2m() */
 	}
 	return true;
 }
@@ -680,14 +655,13 @@ static bool __init early_alloc_p2m_middle(unsigned long pfn)
  * replace the P2M leaf with a p2m_missing or p2m_identity.
  * Stick the old page in the new P2M tree location.
  */
-bool __init early_can_reuse_p2m_middle(unsigned long set_pfn, unsigned long set_mfn)
+static bool __init early_can_reuse_p2m_middle(unsigned long set_pfn)
 {
 	unsigned topidx;
 	unsigned mididx;
 	unsigned ident_pfns;
 	unsigned inv_pfns;
 	unsigned long *p2m;
-	unsigned long *mid_mfn_p;
 	unsigned idx;
 	unsigned long pfn;
 
@@ -733,11 +707,6 @@ bool __init early_can_reuse_p2m_middle(unsigned long set_pfn, unsigned long set_
 found:
 	/* Found one, replace old with p2m_identity or p2m_missing */
 	p2m_top[topidx][mididx] = (ident_pfns ? p2m_identity : p2m_missing);
-	/* And the other for save/restore.. */
-	mid_mfn_p = p2m_top_mfn_p[topidx];
-	/* NOTE: Even if it is a p2m_identity it should still be point to
-	 * a page filled with INVALID_P2M_ENTRY entries. */
-	mid_mfn_p[mididx] = virt_to_mfn(p2m_missing);
 
 	/* Reset where we want to stick the old page in. */
 	topidx = p2m_top_index(set_pfn);
@@ -752,8 +721,6 @@ bool __init early_can_reuse_p2m_middle(unsigned long set_pfn, unsigned long set_
 
 	p2m_init(p2m);
 	p2m_top[topidx][mididx] = p2m;
-	mid_mfn_p = p2m_top_mfn_p[topidx];
-	mid_mfn_p[mididx] = virt_to_mfn(p2m);
 
 	return true;
 }
@@ -763,7 +730,7 @@ bool __init early_set_phys_to_machine(unsigned long pfn, unsigned long mfn)
 		if (!early_alloc_p2m_middle(pfn))
 			return false;
 
-		if (early_can_reuse_p2m_middle(pfn, mfn))
+		if (early_can_reuse_p2m_middle(pfn))
 			return __set_phys_to_machine(pfn, mfn);
 
 		if (!early_alloc_p2m(pfn, false /* boundary crossover OK!*/))

arch/x86/xen/setup.c

@@ -595,6 +595,7 @@ char * __init xen_memory_setup(void)
 		rc = 0;
 	}
 	BUG_ON(rc);
+	BUG_ON(memmap.nr_entries == 0);
 
 	/*
 	 * Xen won't allow a 1:1 mapping to be created to UNUSABLE

arch/x86/xen/time.c

@@ -158,7 +158,7 @@ cycle_t xen_clocksource_read(void)
 	cycle_t ret;
 
 	preempt_disable_notrace();
-	src = this_cpu_ptr(&xen_vcpu->time);
+	src = &__this_cpu_read(xen_vcpu)->time;
 	ret = pvclock_clocksource_read(src);
 	preempt_enable_notrace();
 	return ret;

drivers/xen/balloon.c

@@ -167,6 +167,9 @@ static struct page *balloon_next_page(struct page *page)
 
 static enum bp_state update_schedule(enum bp_state state)
 {
+	if (state == BP_ECANCELED)
+		return BP_ECANCELED;
+
 	if (state == BP_DONE) {
 		balloon_stats.schedule_delay = 1;
 		balloon_stats.retry_count = 1;

drivers/xen/pci.c

@@ -41,24 +41,29 @@ static int xen_add_device(struct device *dev)
 #endif
 
 	if (pci_seg_supported) {
-		struct physdev_pci_device_add add = {
-			.seg = pci_domain_nr(pci_dev->bus),
-			.bus = pci_dev->bus->number,
-			.devfn = pci_dev->devfn
+		struct {
+			struct physdev_pci_device_add add;
+			uint32_t pxm;
+		} add_ext = {
+			.add.seg = pci_domain_nr(pci_dev->bus),
+			.add.bus = pci_dev->bus->number,
+			.add.devfn = pci_dev->devfn
 		};
+		struct physdev_pci_device_add *add = &add_ext.add;
+
 #ifdef CONFIG_ACPI
 		acpi_handle handle;
 #endif
 
 #ifdef CONFIG_PCI_IOV
 		if (pci_dev->is_virtfn) {
-			add.flags = XEN_PCI_DEV_VIRTFN;
-			add.physfn.bus = physfn->bus->number;
-			add.physfn.devfn = physfn->devfn;
+			add->flags = XEN_PCI_DEV_VIRTFN;
+			add->physfn.bus = physfn->bus->number;
+			add->physfn.devfn = physfn->devfn;
 		} else
 #endif
 		if (pci_ari_enabled(pci_dev->bus) && PCI_SLOT(pci_dev->devfn))
-			add.flags = XEN_PCI_DEV_EXTFN;
+			add->flags = XEN_PCI_DEV_EXTFN;
 
 #ifdef CONFIG_ACPI
 		handle = ACPI_HANDLE(&pci_dev->dev);
@@ -77,8 +82,8 @@ static int xen_add_device(struct device *dev)
 				status = acpi_evaluate_integer(handle, "_PXM",
 							       NULL, &pxm);
 				if (ACPI_SUCCESS(status)) {
-					add.optarr[0] = pxm;
-					add.flags |= XEN_PCI_DEV_PXM;
+					add->optarr[0] = pxm;
+					add->flags |= XEN_PCI_DEV_PXM;
 					break;
 				}
 				status = acpi_get_parent(handle, &handle);
@@ -86,7 +91,7 @@ static int xen_add_device(struct device *dev)
 		}
 #endif /* CONFIG_ACPI */
 
-		r = HYPERVISOR_physdev_op(PHYSDEVOP_pci_device_add, &add);
+		r = HYPERVISOR_physdev_op(PHYSDEVOP_pci_device_add, add);
 		if (r != -ENOSYS)
 			return r;
 		pci_seg_supported = false;